Machine Learning Engineer

Agniva Banerjee

E-mail:******@****.***; 848-***-****

Baltimore, Maryland

TECHNICAL SKILLS

● Data Science Libraries: Scikit-learn, Numpy, Pandas, NLTK, SpaCy, CoreNLP, Matplotlib, Seaborn

● Programming Language and Tools: Python, SQL, C#, Azure, Git

● Machine Learning Libraries: PyTorch, Keras, TensorFlow, Fast.ai, fastText, Transformer PROFESSIONAL EXPERIENCE

● Machine Learning Engineer, Leap Orbit LLC, Mar 2018 – Present

● Entity Disambiguation : Predicting whether two entities refer to the same real-world entity

● Worked with stakeholders to define the scope of the problem

● Worked closely with engineers in designing data pipelines needed for downstream use

● Hand crafted shallow models such as Siamese Network and 1D CNN

● Deployed and maintained Machine Learning model in production as API

● Task-specific Information Extraction : Retrieve Task-Specific information from documents

● Used data aggregation and analysis for large semi-structured textual corpus.

● Used standardized pre-processing techniques, performed data cleaning using NLTK

● Used open-source packages for training/fine-tuning deep models such as BERT, ULMFiT

● Software Dev Analyst, Dell International Services, Mar 2014 – Apr 2015

● Envisioned, adaptively planned, and developed an n-tier application (Test Data Portal, used by SIT teams allover Dell IT), with flexible system architecture (.Net, C#).

● Lead a team of 3 developers and quality analysts while developing Test Data Portal.

● Awarded On The Spot Award, December, 2015.

EDUCATION

● University of Maryland at Baltimore County, UMBC Baltimore, Maryland Masters in Computer Science Completed: 05/2018

Select Courses: Machine Learning, Natural Language Processing, Computer Vision RESEARCH EXPERIENCES

PROJECTS:

● Led IBM Cognitive Security project to digest, learn from, build knowledge-graph and reason over vast amounts of structured and unstructured natural language data to quickly discover how to protect against the next zero-day exploit.

● Performed data mining, feature extraction, feature engineering and built shallow neural-networks for predictive modelling tasks on Kaggle datasets as part of course and research-work.

● Led team of 3 on SemEval 2018 Affect in Tweets Task, contributing towards determining the intensity of emotion expressed in tweets on a continuous and discrete scale, by using ensemble learning approach, namely Linear regression, SVM and CNN.

● Developed an intelligent system that analyzes natural language text, extracts relations among entities to find probable product vulnerabilities and recommends the least vulnerable product. GRADUATE ASSISTANTSHIPS

● UMBC Dept. Of Computer Science, Research Assistant for Prof. Tim Finin

● IBM Cognitive CyberSecurity, Language Group

● UMBC Dept. Of Computer Science, Graduate Teaching Assistant

● Artificial Intelligence, Adv. Computer Networks, Introduction to C++ PUBLICATIONS

● Agniva Banerjee, J. C. Martel, “Mitigating the Opioid Epidemic by using Deep Learning toMatch Electronic Health Records”, SIGMOD PODS 2020, under review.

● Dr. Karuna P Joshi, Agniva Banerjee, “Automating Privacy Compliance Using Policy Integrated Blockchain”. Published in: Cryptography Journal as part of the Special Issue Advances of Blockchain Technology and Its Applications, 2018.

● Agniva Banerjee, Dr. Karuna P Joshi, “Link before you Share: Managing Privacy Policies through Blockchain”. Published in: 4th International IEEE PSBD in conjunction with IEEE Big Data, 2017.

● Agniva Banerjee, Raka Dalal, Sudip Mittal, Dr. Karuna P Joshi, “Generating Digital Twin models using Knowledge Graphs for Industrial Production Lines”. Published in: Industrial Knowledge Graphs, 9th International ACM Web-Science Conference, June 2017.

Contact this candidate